Learning Center
Plans & pricing Sign in
Sign Out

Systems biology in the cell nucleus


  • pg 1
									                          Commentary                                                                                                                                     4083

                          Systems biology in the cell nucleus
                          Stanislaw Gorski and Tom Misteli
                          National Cancer Institute, NIH, 41 Library Drive, Bethesda, MD 20892, USA
                          Authors for correspondence (e-mail:;

                          Accepted 20 July 2005
                          Journal of Cell Science 118, 4083-4092 Published by The Company of Biologists 2005

                          The mammalian nucleus is arguably the most complex                                   integrate our knowledge of their sequences and the
                          cellular organelle. It houses the vast majority of an                                molecular mechanisms involved in nuclear processes with
                          organism’s genetic material and is the site of all major                             our insights into the spatial and temporal organization of
                          genome regulatory processes. Reductionist approaches have                            the nucleus and to elucidate the interplay between protein
                          been spectacularly successful at dissecting at the molecular                         and gene networks in regulatory circuits. To do so,
                          level many of the key processes that occur within the                                catalogues of genomes and proteomes as well as a precise
                          nucleus, particularly gene expression. At the same time, the                         understanding of the behavior of molecules in living cells
                          limitations of analyzing single nuclear processes in spatial                         are required. Converging technological developments in
                          and temporal isolation and the validity of generalizing                              genomics, proteomics, dynamics and computation are
                          observations of single gene loci are becoming evident. The                           now leading towards such an integrated biological
                          next level of understanding of genome function is to                                 understanding of genome biology and nuclear function.
Journal of Cell Science

                          Introduction                                                                         organisms do not necessarily use more genes to achieve
                          The sequencing of whole genomes represents a landmark in                             complexity; how gene expression programs are defined in a
                          modern biology (Adams et al., 2000; Goffeau et al., 1996;                            tissue- and cell-type-specific manner; and how the spatial and
                          Lander et al., 2001; Venter et al., 2001; Waterston et al., 2002),                   temporal organization of transcription, RNA processing, RNA
                          revolutionizing the way genes are found, classified and                               degradation, export, DNA replication or DNA repair affects
                          analyzed. It has also brought about a shift in how biological                        genome function. It seems clear now that, apart from genome
                          problems are approached. It has encouraged us to move beyond                         sequence, additional aspects of genome biology must be
                          the limitations of understanding single processes and inspired                       considered if we are to understand how genomes actually work.
                          us to ask how biological events occur at the systems level;                             A key emerging contributor to genome function and
                          rather than analyzing the behavior of a single signaling kinase,                     regulation is the spatial and temporal arrangement of the
                          for example, we now seek to understand how all elements in a                         genome and gene expression processes in nuclear space
                          signaling pathway interact; rather than asking how a gene                            (Misteli, 2001; Spector, 2003). Dramatic developments in
                          responds to an extracellular cue, we now want to know how                            high-resolution and live-cell imaging have revealed that the cell
                          the genome as a whole responds. These types of global                                nucleus is a highly heterogeneous and complex organelle, and
                          approach represent the first steps towards understanding                              that global genome architecture changes during processes such
                          biological functions as they really occur – in the context of                        as differentiation and development (Misteli, 2001; Spector,
                          biological systems.                                                                  2003). One unique feature of the mammalian cell nucleus is
                             While the sequencing of genomes has influenced virtually                           the presence of structural and functional domains that lack
                          all fields of biology, it has had the most profound impact on                         membrane boundaries (Lamond and Earnshaw, 1998; Matera,
                          the study of gene expression itself. This is particularly true                       1999). Spatial organization of the genome is achieved by a
                          because the availability of genome sequence information has                          non-random arrangement of chromosomes in the interphase
                          coincided with the development of microarray analysis, which                         nucleus, with chromosomes occupying preferential
                          allows us to interrogate gene expression at a system level                           intranuclear positions (Cremer and Cremer, 2001; Parada et al.,
                          (Schena et al., 1995). Although most genome-wide analysis                            2004). Chromosomes, genome domains and gene loci may
                          methods are largely descriptive and generally only provide lists                     congress in space to form functional chromatin neighborhoods,
                          of what parts of the genome undergo changes in activity, we                          such as transcriptionally silent heterochromatin regions or
                          can now routinely mine genome-wide expression data, using                            clusters of active genes. In addition to the non-random spatial
                          computational tools to predict biological pathways involved in                       arrangement of genomes, the nucleus also contains numerous
                          a particular physiological response or to group samples, for                         proteinaceous domains, such as the nucleolus and splicing
                          example, normal and malignant tissues, according to                                  factor compartments, which represent distinct structural, and
                          expression patterns (Slonim, 2002). Despite these successes, a                       probably functional, subcompartments (Misteli, 2005).
                          key realization has been that even these comprehensive                               Although the full contribution of the spatial organization of the
                          approaches cannot answer some of the most fundamental                                genome and nuclear proteins is still unknown, it seems clear
                          questions in genome biology, such as why more advanced                               that we must understand genome function in the context of this
                          4084      Journal of Cell Science 118 (18)
                          architectural organization to answer some of the key questions          cellular proteins (Simpson et al., 2000). Although many exist
                          in genome biology.                                                      freely in the nucleoplasm, numerous proteins are enriched in
                             The ultimate goal of a systems biology view of the cell              particular nuclear compartments. A crucial first step towards
                          nucleus is to understand genome function within the                     an integrated view of the nucleus and its compartments is the
                          architectural framework of the nucleus. Gaining such a systems          generation of a protein inventory for each domain (Fig. 2).
                          view of nuclear function will involve several steps. First, we          Proteomic approaches involving biochemical purification and
                          must generate proteomic and genomic inventories of nuclear              mass spectrometry are now beginning to yield this information.
                          components, including genome sequence elements, epigenetic              The first nuclear structure to be analyzed in this manner was
                          modifications, higher-order chromatin structure, chromatin-              the interchromatin granule cluster (IGC), which constitutes the
                          binding complexes and nuclear compartments. Second, we                  major fraction of the splicing factor compartments (Mintz et
                          must understand the cell biological properties of nuclear               al., 1999; Saitoh et al., 2004). Previously, these structures were
                          processes in living cells in terms of characteristics such as their     thought to contain mostly proteins involved in pre-mRNA
                          spatial organization and dynamic properties. Third, we must             splicing, but the detection of >350 proteins, many of which
                          integrate experimental data on process dynamics and spatial             have no apparent function in the splicing reaction, in purified
                          organization, using computational approaches (Fig. 1). Several          IGCs suggests they have additional functions (Mintz et al.,
                          technological developments now make this a realistic, although          1999; Saitoh et al., 2004). Proteomic analysis has now revealed
                          still highly challenging, goal. On the one hand, proteomic and          a similarly astonishing number of proteins associated with the
                          genomic approaches can provide complete lists of components             nucleolus. The latest count for mammalian nucleolar proteins
                          present in particular nuclear structures, and identify what             is ~600 proteins, and the plant nucleolus appears to have at
                          transcription factors are bound where throughout the genome,            least 200 proteins (Andersen et al., 2005; Pendle et al., 2005;
                          and what coding and non-coding genome regions are active at             Scherl et al., 2002). These remarkably high numbers of
                          any given time. On the other hand, quantitative in vivo                 constituents, and the fact that many are novel proteins, provide
                          microscopy methods provide the first glimpse of how DNA,                 a clear indication of how little we know about the functions of
                          RNA and proteins behave inside the nuclei of living cells.              nuclear domains as a whole.
Journal of Cell Science

                          Combining these methods and mining the data by emerging                    Two remarkable commonalities between the proteomic
                          computational approaches will eventually lead to a realistic            analyses of IGCs and the nucleolus are noteworthy. First,
                          picture of gene expression. We review here how far we have              neither revealed the presence of conserved targeting signals in
                          progressed on the long road towards achieving a systems                 resident proteins, which suggests that the enrichment of protein
                          biology view of the cell nucleus and of genome function.                subsets is largely determined by the functions they have in a
                                                                                                  particular compartment rather than by specific targeting
                                                                                                  mechanisms (Dundr and Misteli, 2002). Second, proteins that
                          Nuclear inventories                                                     one would not expect to be present in these compartments are
                            Proteomics in the cell nucleus                                        abundant (Andersen et al., 2005; Saitoh et al., 2004). For
                          The mammalian cell nucleus contains an estimated 20% of all             example, histones are associated with IGCs although these
                                                                                                  compartments are insensitive to DNAse and are generally
                                                                                                  assumed to be chromatin free (Saitoh et al., 2004). Conversely,
                                                                                                  numerous pre-mRNA splicing factors can be detected in the
                                                                                                  nucleolus, despite the fact that the ribosomal genes found
                                                                                                  within the nucleolus do not produce spliceable pre-mRNAs
                                                                                                  (Andersen et al., 2005). Although contaminating proteins may
                                                                                                  contribute to some of these surprising findings, it seems more
                                                                                                  probable that the unexpected appearance of proteins in certain
                                                                                                  compartments reflects their ability to diffuse relatively freely
                                                                                                  throughout the nuclear space and stochastically associate with
                                                                                                  nuclear compartments non-specifically (Dundr and Misteli,
                                                                                                     The behavior of single proteins in response to changes in
                                                                                                  cellular conditions is traditionally followed by in vivo
                                                                                                  microscopy. Proteomic approaches now allow us to track
                                                                                                  changes in the composition of entire subnuclear organelles.
                                                                                                  Proteins can be tagged with multiple metabolic labels, which
                                                                                                  allows us to trace the compositions of entire organelles over
                                                                                                  time (Andersen et al., 2005). When applied to the nucleolus,
                          Fig. 1. Requirements and goals of a systems approach in the nucleus.    this approach demonstrates that, upon inhibition of
                          Systems approaches require the combination of experimental              transcription, nucleolar components can behave very
                          genomic and proteomic information with principles of spatial and        differently: some accumulate in the morphologically altered
                          temporal organization. Computational methods allow hypothesis to
                                                                                                  nucleolus, whereas others dissociate (Andersen et al., 2005).
                          be tested through quantitative predictions that can be verified
                          experimentally. Integration of such data leads to a comprehensive       Remarkably, functionally related proteins or proteins known to
                          view of nuclear processes within the context of nuclear architecture.   be present in a complex, such as pre-ribosomes, often exhibit
                          The results of these interdisciplinary approaches are novel insights    similar kinetics, which suggests that this method can also be
                          into molecular mechanisms and discovery of novel concepts.              used to probe the dynamics of protein complexes and
                                                                                                                   Systems biology in the cell nucleus             4085
                                                                          NUCLEAR PROCESSES
                                                                          • cellular organization (microscopy)
                                                                          • dynamics (microscopy, CHIP)
                          NUCLEAR COMPARTMENTS                            • coordination
                          • composition (microscopy, proteomics)
                          • dynamics (microscopy, proteomics)

                                                                                                                             GENOME STRUCTURE
                                                                                                                             • histone modifications (CHIP)
                                                                                                                             • chromatin structure
                                                           RECRUITMENT                                                       • chromatin dynamics (microscopy)
                                                                                                              ac             • transcription factor dynamics (microscopy)

                                                                                     TF                          ac

                                                                     DYNAMICS                                 meth
                                              SPACIAL                                                         meth

                                           ATGCGT       ATGCGT                                   TF
                                                                   meth     ac meth       ac

Journal of Cell Science

                                                                                                           Fig. 2. Systems information in the cell nucleus. Complete
                                              GENOME ELEMENTS                                              information on genome sequences, DNA-protein interactions,
                                              • sequence elements (sequencing, bioinformatics)             chromatin structure, nuclear compartments and spatial
                                              • nucleosome positions                                       temporal organization is required to develop systems model of
                                              • histone modifications (ChIP-CHIP)                          nuclear function. Experimental and computational tools to
                                              • TF binding sites (Chip)                                    obtain this information are now available.

                          functionally related groups of proteins. These observations               makes them more amenable to purification than nuclear
                          highlight the dynamic nature of the nucleolus, and probably of            compartments. Use of versatile protein-tagging systems now
                          nuclear compartments in general. More importantly, the ability            also allows identification of complexes that are generally more
                          to follow the fate of all components of a nuclear domain under            difficult to purify. A powerful demonstration of such an
                          changing physiological conditions is a powerful tool that will            approach is the recent purification and proteomic analysis of
                          help us to understand how nuclear architecture responds at the            protein complexes that bind to and define the boundary
                          molecular level as a system to signaling cues.                            between silenced chromatin regions at Saccharomyces
                             The components of various other nuclear structures have                cerevisiae telomeres and the neighboring subtelomeric active
                          recently also been cataloged. Several hundred proteins that               regions (Tackett et al., 2005). In this approach, tagged proteins
                          purify with the nuclear envelope have been identified and the              were crosslinked to their binding sites and mapped throughout
                          composition of the nuclear pore complex (NPC) has been                    the genome. Intriguingly, complexes with compositions similar
                          similarly characterized (Rout et al., 2000; Schirmer et al.,              to the one at the telomeric boundary were found at various
                          2003). Since nuclear lamins and associated nuclear envelope               locations throughout the genome, allowing the possibility that
                          proteins have been implicated in various human diseases, this             these sites have similarly, yet unappreciated, boundary
                          former effort has provided leads to numerous novel putative               functions. The ability to analyze not only soluble complexes
                          disease genes and mechanisms (Schirmer et al., 2003). The                 and nuclear compartments but also chromatin-bound large
                          proteomic analysis of the NPC has allowed the spatial mapping             macromolecular assemblies, will be crucial in global mapping
                          of all NPC components within the three-dimensional structure              of protein complexes along the genome.
                          of the pore and, on the basis of homology analysis, functional
                          roles for the various NPC components have been proposed,
                          which can now be experimentally tested (Rout et al., 2003).               Genomics in the cell nucleus
                             Proteomic approaches are not limited to the analysis of                   Beyond coding regions
                          architectural features of the nucleus, but can similarly be               The most basic component of a systems view of the nucleus is
                          applied to nuclear protein complexes. Indeed, the protein                 complete genome expression profiles under various
                          compositions of pre-ribosomes and spliceosomes have been                  physiological conditions (Fig. 2). Although microarray
                          determined (Bassler et al., 2001; Harnpicharnchai et al., 2001;           technology is now a standard tool, most such analyses have
                          Rappsilber et al., 2002; Zhou et al., 2002). Analysis of such             been limited to the <10% of total genome information that
                          complexes is facilitated by their particle-like nature, which             represents coding regions. The recent development of high-
                          4086     Journal of Cell Science 118 (18)
                          density tiling arrays that also contain non-coding regions has         requires the transcription factor binding partner Tec1, but the
                          started to reveal the true complexity of the metazoan                  binding of Ste12p to mating genes does not. Tec1 is negatively
                          transcriptome (Bernstein et al., 2005; Cawley et al., 2004;            regulated by a pheromone-activated MAP kinase, which results
                          Cheng et al., 2005; Kampa et al., 2004). For example, an               in selective binding of Ste12p to mating genes because
                          analysis of ten human chromosomes at 5 bp resolution                   activation of filamentous genes by the Ste12p-Tec1 complex is
                          identified as much as 32% of cellular transcripts to come from          inhibited (Zeitlinger et al., 2003). Therefore, regulation of gene
                          intergenic, non-coding regions, and >60% of the transcripts            expression by common factors may occur through modulation
                          that hybridize with unannotated sequences show transcription           of the activities of their binding partners rather than solely
                          from both genomic strands (Cheng et al., 2005). Systematic             through their interaction with the target sequence.
                          analysis of the function of these non-coding RNAs is                      Genomic approaches have also proven important in
                          particularly important in light of the emerging roles of mirco-        identifying the targets of transcription factors that coordinate
                          RNAs in key physiological processes. These regulatory                  central cellular processes such as proliferation, growth,
                          molecules represent nearly 1% of all genes in higher                   differentiation and apoptosis (Cawley et al., 2004; Fernandez
                          eukaryotes (Bartel, 2004), and have been implicated in                 et al., 2003; Levens, 2003; Orian et al., 2003). For example,
                          developmental timing, neuronal patterning (Lee et al., 1993;           the proto-oncogene Myc has been shown to bind to and
                          Wightman et al., 1993), apoptosis (Brennecke et al., 2003) and         modulate the levels of up to 10% of human genes (Fernandez
                          cellular proliferation (Hatfield et al., 2005), and can be used in      et al., 2003; Orian et al., 2003). A further surprise was the
                          cancer profiling (Lu et al., 2005). These observations make it          location of the binding sites for important transcription factors
                          clear that, just as we are beginning to be able to explore gene        such as Myc, Sp1 and p53 in relation to coding regions
                          expression globally, we must realize that a much larger                (Cawley et al., 2004). Analysis using tiling arrays across
                          percentage of the genome is being transcribed than is presently        chromosomes 21 and 22 showed that these proteins, and
                          appreciated. Determination of the functional impact of                 presumably others, occupy many more sites than previously
                          transcription from these non-coding regions will be crucial to         thought. Moreover, <25% of these binding sites are at the 5
                          our understanding of genome function.                                  end of known genes, but >36% of binding sites are located in
Journal of Cell Science

                                                                                                 or proximal to the 3 regions. Interestingly, many of these sites
                                                                                                 coincide with regions containing non-coding RNAs, which
                             Transcription factor binding                                        suggests the existence of an as-yet-unappreciated level of
                          A further key piece of information is the location of binding          genome regulation by transcription factors through control of
                          sites for the multitude of transcription factors bound along the       non-coding RNA expression (Cawley et al., 2004).
                          genome (Fig. 2). Whereas in-silico bioinformatics has yielded
                          some of this information by mapping consensus sites
                          throughout the genome, approaches involving chromatin                     Chromatin organization and modifications
                          immunoprecipitation (ChIP) coupled with microarray analysis,           One way the usage of a transcription-factor-binding site can be
                          (the so called ChIP-on-Chip approach), can reveal the actual           controlled is by packaging of DNA within chromatin. Indeed,
                          binding sites within the genome for a particular protein under         transcription-factor-binding sites within promoters have
                          a given condition. In ChIP-on-Chip, proteins are crosslinked to        decreased nucleosome occupancy in yeast (Dion et al., 2005;
                          DNA in vivo, they are then immunoprecipitated, the bound               Lee et al., 2004; Yuan et al., 2005). Interestingly, an inverse
                          DNA amplified, and all recovered fragments are identified on             correlation between gene activity and nucleosome occupation
                          microarrays and mapped onto the genome sequence. Such                  is observed when global gene expression is modulated by heat
                          methods are beginning to reveal several surprising mechanistic         shock (Lee et al., 2004). In addition to local DNA accessibility,
                          principles of how proteins interact with the genome.                   higher-order chromatin structure also affects genome function.
                             Bioinformatic analyses generally assume that the presence           Consistent with this view, analysis of a small number of genes
                          of a particular target sequence for a transcription factor means       more than two decades ago suggested that genes that are
                          that it is used; genome-wide analysis by ChIP-on-Chip                  actively transcribed exist in an open conformation, whereas
                          questions this view. In one study, mapping of binding sites for        inactive genes are maintained in a condensed chromatin
                          the transcriptional activator Gal4 showed that not all consensus       structure (Kimura et al., 1983). Recent genome-wide mapping
                          binding sequences are occupied when S. cerevisiae cells are            of open and closed chromatin regions, and their correlation
                          grown in the presence of galactose. This indicates that                with gene activity, paints a more complicated picture. Using
                          sequence itself is not sufficient for a protein to bind (Ren et al.,   fluorescence in situ hybridization and DNA microarray
                          2000). It is not entirely clear what determines whether a              analysis with probes for either open or condensed chromatin
                          specific binding site is occupied, but it seems probable that the       regions, a direct correlation was identified between chromatin
                          combinatorial binding of several proteins as well as the status        compaction and gene density, but not necessarily gene activity
                          of higher-order chromatin structure contributes significantly.          (Gilbert et al., 2004). Gene-rich chromosomes, such as human
                             A concept that has emerged from such analyses is how a              chromosome 22, contain large regions of chromatin in an open
                          single factor can regulate the expression of different sets of         conformation compared with gene-poor regions within
                          genes. This behavior is exemplified by the transcription factor         chromosome 1. Despite this structural heterogeneity, a good
                          Ste12, in S. cerevisiae. Upon the addition of pheromone,               correlation with gene transcription was not observed because
                          Ste12p activates genes involved in mating, whereas during              many silenced genes are also located in these open chromatin
                          starvation the same factor activates genes in the filamentous           regions (Gilbert et al., 2004). These results suggest that the role
                          growth pathway. Genome-wide location analysis showed that              of chromatin in the regulation of gene expression occurs on a
                          binding of Ste12p to genes that regulate filamentous growth             more local basis. Such a notion is consistent with results from
                                                                                                          Systems biology in the cell nucleus            4087
                          genome-wide binding studies demonstrating that chromatin-            early, and some portions of euchromatic chromosome regions
                          remodeling and -modifying complexes that can change the              also replicate late in S phase (Schubeler et al., 2002).
                          local structure of chromatin and expose binding sites are            Intriguingly, tiling arrays have shown that highly transcribed
                          recruited to promoters of many genes upon activation (Lusser         sequences outside annotated genes also replicate early during
                          and Kadonaga, 2003; Robert et al., 2004).                            S phase (White et al., 2004) and that there is a relationship
                             The functional status of chromatin often broadly correlates       between gene density and replication timing, most dense
                          with post-translational modifications of the core histones and        regions replicating in the first 4 hours of S phase (Jeon et al.,
                          these modifications have been proposed to constitute an               2005). These first glimpses into the global relationship between
                          epigenetic code (Strahl and Allis, 2000). Genomic analyses can       replication timing, gene activity, gene organization and
                          identify the patterns of these modifications and link them to         chromatin status make it clear that these are complex links that
                          features of the genome such as regulatory sequences and to           are difficult to understand through analysis of single loci, and
                          gene activity (Fig. 2) (Bernstein et al., 2005; Dion et al., 2005;   that the regulatory mechanisms and the precise nature of the
                          Kurdistani et al., 2004; Roh et al., 2005; Schubeler et al.,         local chromatin structure at these anomalous regions remain to
                          2004). For example, in human T cells the distribution and            be discovered.
                          relative quantities of acetylated histones H3 and H4 have been
                          mapped across the genome (Roh et al., 2005). This method was
                          able to identify islands of acetylation that preferentially occur    Dynamics
                          at known promoters and regulatory sequences rather than in the       Genome-wide mapping of sites at which proteins interact with
                          main body of genes.                                                  chromatin provides a detailed picture of the steady-state
                             Similar studies are beginning to reveal patterns of different     distribution of proteins within the genome at a specific time.
                          modifications in relation to gene activity and organization           However, temporal aspects of their behavior add an additional
                          (Bernstein et al., 2005; Schubeler et al., 2004). In yeast, for      level of complexity that must be taken into account (Fig. 2).
                          example, histone acetylation has a combinatorial effect              Time-lapse      microscopy       experiments     coupled     with
                          (Kurdistani et al., 2004). When the distribution of the              photobleaching protocols have demonstrated that most
Journal of Cell Science

                          acetylation status of eleven different lysines in the same histone   chromatin proteins can move freely and rapidly through the
                          octamer region was determined, distinct combinations of              nuclear space by passive diffusion and, more importantly, that
                          acetylated and deacetylated lysines were observed at specific         most protein-chromatin interactions are transient, including
                          genomic locations. Intriguingly, gene regions that have similar      those involving structural components of chromatin (Agresti et
                          modification patterns are co-regulated, which supports the            al., 2005; Cheutin et al., 2003; Festenstein et al., 2003; Kimura
                          notion of a combinatorial code (Kurdistani et al., 2004).            and Cook, 2001; Lever et al., 2000; McNally et al., 2000;
                          However, the specific roles these acetylation patterns play in        Misteli, 2001; Misteli et al., 2000; Phair et al., 2004b). The
                          recruiting histone-associated proteins are still not clear. It has   rapid exchange kinetics – typical residence times of proteins
                          been a long standing view that acetylation of histones could         on chromatin are of the order of seconds – suggests that
                          play a non-specific role in maintenance of chromatin structure        chromatin is highly dynamic (Bustin et al., 2005). This
                          and dynamics (Horn and Peterson, 2002). Indeed, an analysis          property is probably essential for the genome to respond
                          of combinations of four different modifications in the same           rapidly to changing environmental conditions. It has
                          octamer unit, demonstrated that three of the residues do not         implications for how proteins interact with the genome
                          play a specific role in transcription but pointed towards a           globally as well as locally and thus for the system-level
                          general cumulative effect, potentially on chromatin structure        behavior of the nucleus and the genome.
                          (Dion et al., 2005). Despite these differences, the power of a          The ability of proteins to move rapidly throughout the
                          system-wide analysis of histone modifications should lead to          nucleus ensures their availability throughout the genome
                          elucidation of the functional link between these modifications        (Misteli, 2001). In addition, the short residence times on
                          and gene activity.                                                   chromatin allow proteins to search the genome for specific
                                                                                               binding sites by three-dimensional scanning as a molecule
                                                                                               briefly binds to chromatin, dissociates and rapidly binds to
                             Systematic mapping of DNA replication sites                       another site (Misteli, 2001). Upon a chance encounter with one
                          The structural state of chromatin has also been implicated in        of its specific sites, the molecule resides for a longer period of
                          nuclear functions other than gene expression. There is a long-       time and exerts its full biological activity. Some molecules may
                          recognized link between replication timing and chromatin             also function at non-specific binding sites. For example,
                          compaction: frequently euchromatic regions replicate early           chromatin-remodeling factors may continuously remodel
                          whereas heterochromatic regions replicate later during S phase       chromatin stochastically, regardless of the nature of the binding
                          (Gilbert, 2002). Genome-wide analysis of sequence-dependent          site, and in that way contribute significantly to the overall
                          replication timing in higher organisms such as Drosophila and        dynamic nature of chromatin.
                          humans has begun to reveal some relationships between                   At the local level, dynamic binding significantly affects the
                          genome structure and activity, and added more levels of              behavior of proteins because it provides the basis for regulatory
                          complexity (Jeon et al., 2005; Schubeler et al., 2002; White et      events through combinatorial, competitive networks of
                          al., 2004; Woodfine et al., 2004). As previously appreciated,         interactions. The transient binding of a given protein means
                          for many early-replicating genes replication timing correlates       that each time a molecule dissociates from its binding site, the
                          with their transcriptional activity, and most repetitive             vacant site can be occupied by another protein. Since this might
                          heterochromatic regions replicate late during S phase.               have a distinct function, the fate of the region depends on the
                          However, other proximal heterochromatic regions replicate            competitive binding of these factors. An example of this is the
                          4088     Journal of Cell Science 118 (18)
                          binding of the linker histone H1 and the high mobility group        made quantitative predictions. Furthermore, it illustrates that
                          (HMG) proteins (Bustin et al., 2005; Catez et al., 2002). H1 is     the behavior of complex processes such as nuclear import is
                          generally considered a transcriptional repressor, whereas HMG       difficult to predict intuitively and that quantitative models are
                          proteins generally act as activators. The binding sites for         required to do so.
                          several HMG proteins overlap with that of H1, and                      Cao and Parker have similarly combined quantitative
                          experimental titration of HMG into living cells results in          experimental data and in-silico modeling to analyze normal
                          increased occupancy of DNA by HMG proteins through                  mRNA-turnover and nonsense-mediated decay in S. cerevisiae
                          competition for the same binding site (Catez et al., 2004).         (Cao and Parker, 2001; Cao and Parker, 2003). Apart from
                          Similarly, the relevance of dynamic interaction networks has        generating estimates of quantitative values for key steps in the
                          become apparent from a systematic analysis of the assembly          degradation process such as de-adenylation and de-capping,
                          dynamics of the RNA polymerase I transcription machinery            they also used the model to test the hypothesis that preferential
                          with endogenous ribosomal genes (Dundr et al., 2002). GFP           degradation of RNAs containing nonsense codons near the 5
                          fusion proteins that contain pre-initiation, assembly factors and   end is due to leaky mRNA surveillance in which the
                          different polymerase I subunits have distinct binding kinetics,     recognition of nonsense codons near the 5 end of the mRNA
                          which suggests that their interactions at the promoter occur        is more efficient than recognition of a 3 nonsense codon
                          individually as part of a dynamic interaction network. It will      (Ishigaki et al., 2001). However, their model reveals that the
                          be important to determine whether changes in the interaction        leaky surveillance cannot account for the observed degradation
                          dynamics are responsible for changes in the transcriptional         kinetics (Cao and Parker, 2003). Rather, the experimental data
                          output of the polymerase.                                           can only be explained by a mechanism in which all nonsense-
                                                                                              codon-containing mRNAs are recognized equally well but their
                                                                                              decapping and subsequent degradation rates differ. This
                          Computation                                                         quantitative analysis of NMD has now given rise to a new,
                          Analysis of complex biological processes in the cell nucleus        testable model of NMD referred to as position-independent
                          requires the use of computational tools. In addition to the well-   efficient surveillance (Cao and Parker, 2003).
Journal of Cell Science

                          established methods for mining gene expression data to                 This type of computational analysis has also been applied to
                          identify gene networks and pathways, simulation is becoming         the assembly and elongation behavior of a transcription
                          indispensable for the analysis of the kinetics of nuclear           complex (Dundr et al., 2002). Dundr et al. fluorescently tagged
                          processes (Phair et al., 2004a; Slepchenko et al., 2002; Sprague    the majority of the subunits of the RNA polymerase I complex
                          and McNally, 2005). In these approaches, experimental data          and analysed their association and elongation kinetics
                          are used to determine numerical values for kinetic parameters       on endogenous ribosomal genes, using photobleaching
                          and these parameters then serve as constraints in computational     microscopy. This allowed them to estimate the in vivo RNA
                          models that describe a particular process (Phair et al., 2004a;     polymerase I elongation rate and show that the subunit
                          Slepchenko et al., 2002; Sprague and McNally, 2005). Such           behavior is consistent with a model in which the polymerase
                          models, which often consist of systems of differential              assembles in a stepwise fashion from its subunits via formation
                          equations, can then be used to make quantitative predictions,       of metastable intermediates (Dundr et al., 2002).
                          which in turn can be tested experimentally. Not only does this         Many nuclear processes are tightly interconnected (Maniatis
                          iterative process provide quantitative information and              and Reed, 2002; Neugebauer, 2002). For example, pre-mRNA
                          predictions, but the failure of a given model to account for        processing steps, including poly-adenylation and splicing,
                          experimentally observed properties often points to novel            occur co-transcriptionally, and the RNA-processing machinery
                          conceptual aspects of a process. These types of computational       is physically associated with elongating transcription
                          approach are now being applied to the analysis of nuclear           complexes. Similarly, the trigger for RNA export is intimately
                          processes.                                                          linked to completion of pre-mRNA splicing (Tange et al.,
                             The most advanced system-level analysis of any nuclear           2004), and even the choice between alternative splice sites is
                          function is probably the in-silico analysis of nuclear protein      influenced by transcriptional co-activators and promoter usage
                          import (Gorlich et al., 2003; Riddick and Macara, 2005; Smith       (Cramer et al., 1997). The next step in computational
                          et al., 2002). Here, the authors microinjected cells with           modeling, therefore, will be to interrogate combinations of
                          fluorescently labeled cargo and measured nuclear import rates        multiple processes. This poses an experimental challenge
                          under various experimental conditions, as well as using in-vitro    because, ideally, information about multiple steps in the gene
                          analysis to determine several crucial parameters such as on-        expression process should be extracted from a single
                          and off-rates for various import intermediates. Developing          experimental system. This is rarely possible, particularly in the
                          computational models of the import pathway, they were able          context of nuclear architecture. However, Janicki et al. have
                          to examine the contribution of the import-driving Ran gradient      developed a potentially powerful tool that holds much promise
                          as well as that of importin (Gorlich et al., 2003; Riddick and      for analyzing sequential steps from transcription to protein
                          Macara, 2005; Smith et al., 2002). This system demonstrated         synthesis (Janicki et al., 2004). They have generated an
                          the value of modeling approaches because it generates several       artificial gene array that allows visualization of changes in
                          counterintuitive predictions, including a limiting role of          chromatin structure, binding of transcription factors, mRNA
                          importin and an inhibitory role of excess importin – both           synthesis, RNA export and protein synthesis. This system has
                          of which were confirmed when tested experimentally (Riddick          already been used to quantitatively track the movement of
                          and Macara, 2005). The model for nuclear import has thus not        single mRNA-protein complexes (mRNPs) in living cells,
                          only provided the first quantitative assessment of many of the       demonstrating that mRNAs move rapidly by diffusion through
                          import steps but also given novel mechanistic insights and          the nucleus (Shav-Tal et al., 2004). Notwithstanding the caveat
                                                                                                              Systems biology in the cell nucleus           4089
                          of an artificial experimental system, the application of                  2004). It has also become clear that specific repair pathways
                          photobleaching methods in combination with computational                 share common members and even functions. An example of
                          analysis should provide a comprehensive quantitative picture             this is the base-excision-repair pathway, which is linked to the
                          of the coordination of gene expression steps.                            nucleotide excision, mismatch and recombinatorial repair
                                                                                                   pathways through various molecular interactions (Begley and
                                                                                                   Samson, 2004).
                          Towards an integrated view of nuclear function                              An extension of the traditional transcription profiling
                          The ultimate goal of systems biology in the nucleus is a                 approach is the combination of genome-wide transcription-
                          computational model that is able to describe all nuclear                 factor-binding information with bioinformatics to mine
                          processes from genome replication, via maintenance of DNA                sequence elements in yeast (Harbison et al., 2004). This
                          integrity, to regulation of expression within the spatial and            approach generates a molecular network of all potential
                          temporal framework of nuclear architecture. If this were not             transcription-factor-binding sites and motifs under various
                          ambitious enough, the benchmark for such a model will be its             environmental conditions. Integrating the data on different
                          ability to generate quantitative predictions in silico that can be       transcription-factor-binding sites produces a ‘transcriptional-
                          tested experimentally. Achieving this goal will not be possible          potential-network’ that provides a framework for modeling the
                          in a single step but will involve the development of gradually           mechanisms that lead to global gene expression (Harbison et
                          more complex, and presumably more accurate, computational                al., 2004). Analyzing data in this manner is important because
                          models by iterative comparison of experimental and simulation            it allows us to identify genes that share similar combinations
                          data. Two complementary strategies should eventually achieve             of transcription-factor-binding sites and regulatory motifs, and
                          this goal (Fig. 3).                                                      draw general conclusions concerning organization of gene
                             Genome and proteome information can be directly                       sequences.
                          assembled into pathways and networks (Fig. 3). In these                     Although this type of network modeling provides insights
                          approaches, little attention is paid to cellular organization or         into the overall properties of physiological processes, it is
                          the details of molecular interactions, but only functional               relatively low-resolution and generally ignores most aspects of
Journal of Cell Science

                          relationships between pathway components are considered. A               how biological processes are organized in cells. To obtain a
                          good example of this approach is the analysis of DNA-damage              more realistic, complete system-level view we must model
                          responses using microarray methods (Begley and Samson,                   single processes in the greatest possible detail, taking into
                          2004) Analysis of global changes in gene expression shows                account the networks of molecular interactions, their dynamics
                          that expression patterns of genes involved in a wide variety of          and spatial organization (Fig. 3). Processes that have been
                          processes, such as protein degradation and synthesis, signal             explored in this way include transcription by RNA polymerase
                          transduction, RNA metabolism and chromatin remodeling as                 I machinery, nuclear import and NF- B signaling (Dundr et al.,
                          well as transcription and DNA repair, are altered upon                   2002; Hoffmann et al., 2002; Nelson et al., 2004; Riddick and
                          treatment with DNA damage reagents (Begley and Samson,                   Macara, 2005; Smith et al., 2002). Each can be considered a
                                                                                                   functional module in a larger network, and ultimately the
                                                                                                   computational models for multiple such functional modules
                                                                                                   will be combined to give a comprehensive global view of
                                                                                                   nuclear function.
                                                                                                      This type of bottom-up approach to systems biology is more
                                                                                                   difficult owing to the incorporation of more detail, but at the
                                                                                                   same time it has several important advantages. Analysis of
                                                                                                   limited functional modules allows development of relatively
                                                                                                   simple computational models that can be tested experimentally
                                                                                                   and are easier to grasp than large networks. For example, in
                                                                                                   the case of NF- B signaling, Hoffmann et al. analyzed the
                                                                                                   effects of knocking-out each known inhibitor of NF- B
                                                                                                   (Hoffmann et al., 2002). Quantitation of the response to
                                                                                                   stimulation and computational modeling identified specific
                                                                                                   roles for each inhibitor, and provided a global, yet detailed,
                                                                                                   view of how this particular pathway behaves (Hoffmann et al.,
                                                                                                   2002). In addition, modeling of distinct functional modules
                                                                                                   using quantitative imaging combined with computational
                                                                                                   analysis allows the incorporation of spatial and temporal
                                                                                                   information, i.e. the role of subcellular localization, local
                                                                                                   protein concentrations or wave-like propagation of a signal. In
                          Fig. 3. Strategies for the development of comprehensive system           the case of NF- B, quantitative live-cell imaging has shown
                          models. Proteomic or genomic information can be directly used to
                                                                                                   that persistent asynchronous oscillations between the nucleus
                          develop pathway and network models (dashed lines). Alternatively,
                          specific cellular processes such as transcription or replication can be   and cytoplasm are required for NF- B target gene expression.
                          modeled in the form of functional modules, using quantitative cell-      The authors suggest, that akin to Ca2+ signaling, the
                          biological methods including combined imaging and computational          consequences of NF- B may depend on the number, period
                          approaches. Linking models of functional modules yields an               and amplitude of oscillations (Nelson et al., 2004). Such cell
                          integrated model.                                                        biological properties are often ignored in network modeling
                          4090     Journal of Cell Science 118 (18)
                          approaches, despite their established importance in regulation        References
                          of biological processes.                                              Adams, M. D., Celniker, S. E., Holt, R. A., Evans, C. A., Gocayne, J. D.,
                             Once single functional modules are satisfactorily simulated,         Amanatides, P. G., Scherer, S. E., Li, P. W., Hoskins, R. A., Galle, R. F.
                                                                                                  et al. (2000). The genome sequence of Drosophila melanogaster. Science
                          they can be linked into larger process networks. An example             287, 2185-2195.
                          of this approach is a system-level view of sea urchin                 Agresti, A., Scaffidi, P., Riva, A., Caiolfa, V. R. and Bianchi, M. E. (2005).
                          development. Davidson and co-workers first analyzed the                  GR and HMGB1 interact only within chromatin and influence each other’s
                          spatial and temporal regulation of genes responsible for cell           residence time. Mol. Cell 18, 109-121.
                          specification, using a bottom-up modular approach (Yuh et              Andersen, J. S., Lam, Y. W., Leung, A. K., Ong, S. E., Lyon, C. E.,
                                                                                                  Lamond, A. I. and Mann, M. (2005). Nucleolar proteome dynamics.
                          al., 1998). They combined quantitative experiments and                  Nature 433, 77-83.
                          computational analysis to identify the roles of different cis-        Bartel, D. P. (2004). MicroRNAs: genomics, biogenesis, mechanism, and
                          regulatory sequences controlling expression of a protein                function. Cell 116, 281-297.
                          involved in the development of the midgut (Yuh et al., 1998;          Bassler, J., Grandi, P., Gadal, O., Lessmann, T., Petfalski, E., Tollervey,
                                                                                                  D., Lechner, J. and Hurt, E. (2001). Identification of a 60S preribosomal
                          Yuh et al., 2001). They then verified this approach by                   particle that is closely linked to nuclear export. Mol. Cell 8, 517-529.
                          confirming in-silico-generated predictions experimentally              Begley, T. J. and Samson, L. D. (2004). Network responses to DNA damaging
                          (Yuh et al., 1998). The subsequent addition of more modules             agents. DNA Repair 3, 1123-1132.
                          representing different spatial domains has allowed them to            Bentele, M., Lavrik, I., Ulrich, M., Stosser, S., Heermann, D. W., Kalthoff,
                          assemble a large regulatory network of the genes and                    H., Krammer, P. H. and Eils, R. (2004). Mathematical modeling reveals
                                                                                                  threshold mechanism in CD95-induced apoptosis. J. Cell Biol. 166, 839-
                          regulatory sequences involved in endomesoderm development               851.
                          (Davidson et al., 2002a; Davidson et al., 2002b). Furthermore,        Bernstein, B. E., Kamal, M., Lindblad-Toh, K., Bekiranov, S., Bailey, D.
                          they can now perform large-scale system perturbations and               K., Huebert, D. J., McMahon, S., Karlsson, E. K., Kulbokas, E. J., 3rd,
                          quantitatively describe the resulting changes in multiple               Gingeras, T. R. et al. (2005). Genomic maps and comparative analysis of
                                                                                                  histone modifications in human and mouse. Cell 120, 169-181.
                          phenotypic parameters (Davidson et al., 2002a; Davidson et al.,       Brennecke, J., Hipfner, D. R., Stark, A., Russell, R. B. and Cohen, S. M.
                          2002b).                                                                 (2003). bantam encodes a developmentally regulated microRNA that
                             The top-down and bottom-up modular strategies described              controls cell proliferation and regulates the proapoptotic gene hid in
Journal of Cell Science

                          above are not mutually exclusive but highly complementary.              Drosophila. Cell 113, 25-36.
                                                                                                Bustin, M., Catez, F. and Lim, J. H. (2005). The dynamics of histone H1
                          An impressive example in which they have been combined is               function in chromatin. Mol. Cell 17, 617-620.
                          computational analysis of apoptosis. In this case Bentele et al.      Cao, D. and Parker, R. (2001). Computational modeling of eukaryotic mRNA
                          modeled the well-studied parts of the network in great detail,          turnover. RNA 7, 1192-1212.
                          using experimentally determined rate constants, whereas they          Cao, D. and Parker, R. (2003). Computational modeling and experimental
                          modeled the less well-studied parts at low resolution, only             analysis of nonsense-mediated decay in yeast. Cell 113, 533-545.
                                                                                                Catez, F., Brown, D. T., Misteli, T. and Bustin, M. (2002). Competition
                          taking into account the overall behavior of particular modules          between histone H1 and HMGN proteins for chromatin binding sites. EMBO
                          (Bentele et al., 2004).                                                 Rep. 3, 760-766.
                                                                                                Catez, F., Yang, H., Tracey, K. J., Reeves, R., Misteli, T. and Bustin, M.
                                                                                                  (2004). Network of dynamic interactions between histone H1 and high-
                          Conclusion                                                              mobility-group proteins in chromatin. Mol. Cell. Biol. 24, 4321-4628.
                                                                                                Cawley, S., Bekiranov, S., Ng, H. H., Kapranov, P., Sekinger, E. A., Kampa,
                          How close are we to a system-level view of the nucleus?                 D., Piccolboni, A., Sementchenko, V., Cheng, J., Williams, A. J. et al.
                          Clearly we are still at the very beginning, but the tools that will     (2004). Unbiased mapping of transcription factor binding sites along human
                          make it possible to explore the nucleus on a more global scale          chromosomes 21 and 22 points to widespread regulation of noncoding
                          are being put in place, and the first examples of large scale,           RNAs. Cell 116, 499-509.
                                                                                                Cheng, J., Kapranov, P., Drenkow, J., Dike, S., Brubaker, S., Patel, S.,
                          complex systems analyses are emerging. The relatively                   Long, J., Stern, D., Tammana, H., Helt, G. et al. (2005). Transcriptional
                          simplistic computational models that have been applied have             maps of 10 human chromosomes at 5-nucleotide resolution. Science 308,
                          already demonstrated that quantitative systems analysis has the         1149-1154.
                          potential to provide new insights into the molecular                  Cheutin, T., McNairn, A. J., Jenuwein, T., Gilbert, D. M., Singh, P. B. and
                                                                                                  Misteli, T. (2003). Maintenance of stable heterochromatin domains by
                          mechanisms of nuclear process and also to generate novel                dynamic HP1 binding. Science 299, 721-725.
                          concepts of how cellular processes are organized, coordinated         Cramer, P., Pesce, C. G., Baralle, F. E. and Kornblihtt, A. R. (1997).
                          and regulated. In the future we will probably see an increasing         Functional association between promoter structure and transcript alternative
                          convergence of experimental and in-silico analysis, both of             splicing. Proc. Natl. Acad. Sci. USA 94, 11456-11460.
                                                                                                Cremer, T. and Cremer, C. (2001). Chromosome territories, nuclear
                          particular cellular processes as well as of process networks.           architecture and gene regulation in mammalian cells. Nat. Rev. Genet. 2,
                          These approaches are conceptually different from how                    292-301.
                          biochemists and molecular biologist have traditionally thought        Davidson, E. H., Rast, J. P., Oliveri, P., Ransick, A., Calestani, C., Yuh, C.
                          about problems in the past and will require us to embrace               H., Minokawa, T., Amore, G., Hinman, V., Arenas-Mena, C. et al.
                          multi-disciplinary approaches. While this would have been a             (2002a). A genomic regulatory network for development. Science 295,
                          serious hurdle only a few years ago, fortunately the sequencing       Davidson, E. H., Rast, J. P., Oliveri, P., Ransick, A., Calestani, C., Yuh, C.
                          of whole genomes has provided the experimental foundation,              H., Minokawa, T., Amore, G., Hinman, V., Arenas-Mena, C. et al.
                          as well as the inspiration, for us to tackle cell biological            (2002b). A provisional regulatory gene network for specification of
                          problems, be it in the cell nucleus or elsewhere, at the systems        endomesoderm in the sea urchin embryo. Dev. Biol. 246, 162-190.
                                                                                                Dion, M. F., Altschuler, S. J., Wu, L. F. and Rando, O. J. (2005). Genomic
                          level.                                                                  characterization reveals a simple histone H4 acetylation code. Proc. Natl.
                                                                                                  Acad. Sci. USA 102, 5501-5506.
                            This research was supported by the Intramural Research Program      Dundr, M. and Misteli, T. (2002). Nucleolomics: an inventory of the
                          of the NIH, National Cancer Institute, Center for Cancer Research.      nucleolus. Mol. Cell 9, 5-7.
                          T.M. is a fellow of the Keith R. Porter Endowment for Cell Biology.   Dundr, M., Hoffmann-Rohrer, U., Hu, Q., Grummt, I., Rothblum, L. I.,
                                                                                                                         Systems biology in the cell nucleus                        4091
                            Phair, R. D. and Misteli, T. (2002). A kinetic framework for a mammalian        Lu, J., Getz, G., Miska, E. A., Alvarez-Saavedra, E., Lamb, J., Peck, D.,
                            RNA polymerase in vivo. Science 298, 1623-1626.                                   Sweet-Cordero, A., Ebert, B. L., Mak, R. H., Ferrando, A. A. et al.
                          Fernandez, P. C., Frank, S. R., Wang, L., Schroeder, M., Liu, S., Greene,           (2005). MicroRNA expression profiles classify human cancers. Nature 435,
                            J., Cocito, A. and Amati, B. (2003). Genomic targets of the human c-Myc           834-838.
                            protein. Genes Dev. 17, 1115-1129.                                              Lusser, A. and Kadonaga, J. T. (2003). Chromatin remodeling by ATP-
                          Festenstein, R., Pagakis, S. N., Hiragami, K., Lyon, D., Verreault, A.,             dependent molecular machines. BioEssays 25, 1192-1200.
                            Sekkali, B. and Kioussis, D. (2003). Modulation of heterochromatin              Maniatis, T. and Reed, R. (2002). An extensive network of coupling among
                            protein 1 dynamics in primary Mammalian cells. Science 299, 719-721.              gene expression machines. Nature 416, 499-506.
                          Gilbert, D. M. (2002). Replication timing and transcriptional control: beyond     Matera, A. G. (1999). Nuclear bodies: multifaceted subdomains of the
                            cause and effect. Curr. Opin. Cell Biol. 14, 377-383.                             interchromatin space. Trends Cell. Biol. 9, 302-309.
                          Gilbert, N., Boyle, S., Fiegler, H., Woodfine, K., Carter, N. P. and               McNally, J. G., Muller, W. G., Walker, D., Wolford, R. and Hager, G. L.
                            Bickmore, W. A. (2004). Chromatin architecture of the human genome:               (2000). The glucocorticoid receptor: rapid exchange with regulatory sites in
                            gene-rich domains are enriched in open chromatin fibers. Cell 118, 555-566.        living cells. Science 287, 1262-1265.
                          Goffeau, A., Barrell, B. G., Bussey, H., Davis, R. W., Dujon, B., Feldmann,       Mintz, P. J., Patterson, S. D., Neuwald, A. F., Spahr, C. S. and Spector, D.
                            H., Galibert, F., Hoheisel, J. D., Jacq, C., Johnston, M. et al. (1996). Life     L. (1999). Purification and biochemical characterization of interchromatin
                            with 6000 genes. Science 274, 546, 563-567.                                       granule clusters. EMBO J. 18, 4308-4320.
                          Gorlich, D., Seewald, M. J. and Ribbeck, K. (2003). Characterization of           Misteli, T. (2001). Protein dynamics: implications for nuclear architecture and
                            Ran-driven cargo transport and the RanGTPase system by kinetic                    gene expression. Science 291, 843-847.
                            measurements and computer simulation. EMBO J. 22, 1088-1100.                    Misteli, T. (2005). Concepts in nuclear architecture. BioEssays 27, 477-487.
                          Harbison, C. T., Gordon, D. B., Lee, T. I., Rinaldi, N. J., Macisaac, K. D.,      Misteli, T., Gunjan, A., Hock, R., Bustin, M. and Brown, D. T. (2000).
                            Danford, T. W., Hannett, N. M., Tagne, J. B., Reynolds, D. B., Yoo, J. et         Dynamic binding of histone H1 to chromatin in living cells. Nature 408,
                            al. (2004). Transcriptional regulatory code of a eukaryotic genome. Nature        877-881.
                            431, 99-104.                                                                    Nelson, D. E., Ihekwaba, A. E., Elliott, M., Johnson, J. R., Gibney, C. A.,
                          Harnpicharnchai, P., Jakovljevic, J., Horsey, E., Miles, T., Roman, J.,             Foreman, B. E., Nelson, G., See, V., Horton, C. A., Spiller, D. G. et al.
                            Rout, M., Meagher, D., Imai, B., Guo, Y., Brame, C. J. et al. (2001).             (2004). Oscillations in NF-kappaB signaling control the dynamics of gene
                            Composition and functional characterization of yeast 66S ribosome                 expression. Science 306, 704-708.
                            assembly intermediates. Mol. Cell 8, 505-515.                                   Neugebauer, K. M. (2002). On the importance of being co-transcriptional. J.
                          Hatfield, S. D., Shcherbata, H. R., Fischer, K. A., Nakahara, K., Carthew,           Cell Sci. 115, 3865-3871.
                            R. W. and Ruohola-Baker, H. (2005). Stem cell division is regulated by          Orian, A., van Steensel, B., Delrow, J., Bussemaker, H. J., Li, L., Sawado,
Journal of Cell Science

                            the microRNA pathway. Nature 435, 974-978.                                        T., Williams, E., Loo, L. W., Cowley, S. M., Yost, C. et al. (2003).
                          Hoffmann, A., Levchenko, A., Scott, M. L. and Baltimore, D. (2002). The             Genomic binding by the Drosophila Myc, Max, Mad/Mnt transcription
                            IkappaB-NF-kappaB signaling module: temporal control and selective gene           factor network. Genes Dev. 17, 1101-1114.
                            activation. Science 298, 1241-1245.                                             Parada, L. A., Sotiriou, S. and Misteli, T. (2004). Spatial genome
                          Horn, P. J. and Peterson, C. L. (2002). Molecular biology. Chromatin higher         organization. Exp. Cell Res. 296, 64-70.
                            order folding–wrapping up transcription. Science 297, 1824-1827.                Pendle, A. F., Clark, G. P., Boon, R., Lewandowska, D., Lam, Y. W.,
                          Ishigaki, Y., Li, X., Serin, G. and Maquat, L. E. (2001). Evidence for a            Andersen, J., Mann, M., Lamond, A. I., Brown, J. W. and Shaw, P. J.
                            pioneer round of mRNA translation: mRNAs subject to nonsense-mediated             (2005). Proteomic analysis of the Arabidopsis nucleolus suggests novel
                            decay in mammalian cells are bound by CBP80 and CBP20. Cell 106, 607-             nucleolar functions. Mol. Biol. Cell 16, 260-269.
                            617.                                                                            Phair, R. D., Gorski, S. A. and Misteli, T. (2004a). Measurement of dynamic
                          Janicki, S. M., Tsukamoto, T., Salghetti, S. E., Tansey, W. P.,                     protein binding to chromatin in vivo, using photobleaching microscopy.
                            Sachidanandam, R., Prasanth, K. V., Ried, T., Shav-Tal, Y., Bertrand,             Methods Enzymol. 375, 393-414.
                            E., Singer, R. H. et al. (2004). From silencing to gene expression: real-time   Phair, R. D., Scaffidi, P., Elbi, C., Vecerova, J., Dey, A., Ozato, K., Brown,
                            analysis in single cells. Cell 116, 683-698.                                      D. T., Hager, G., Bustin, M. and Misteli, T. (2004b). Global nature of
                          Jeon, Y., Bekiranov, S., Karnani, N., Kapranov, P., Ghosh, S., Macalpine,           dynamic protein-chromatin interactions in vivo: three-dimensional genome
                            D., Lee, C., Hwang, D. S., Gingeras, T. R. and Dutta, A. (2005). Temporal         scanning and dynamic interaction networks of chromatin proteins. Mol.
                            profile of replication of human chromosomes. Proc. Natl. Acad. Sci. USA            Cell. Biol. 24, 6393-6402.
                            102, 6419-6424.                                                                 Rappsilber, J., Ryder, U., Lamond, A. I. and Mann, M. (2002). Large-scale
                          Kampa, D., Cheng, J., Kapranov, P., Yamanaka, M., Brubaker, S., Cawley,             proteomic analysis of the human spliceosome. Genome Res. 12, 1231-1245.
                            S., Drenkow, J., Piccolboni, A., Bekiranov, S., Helt, G. et al. (2004).         Ren, B., Robert, F., Wyrick, J. J., Aparicio, O., Jennings, E. G., Simon, I.,
                            Novel RNAs identified from an in-depth analysis of the transcriptome of            Zeitlinger, J., Schreiber, J., Hannett, N., Kanin, E. et al. (2000). Genome-
                            human chromosomes 21 and 22. Genome Res. 14, 331-342.                             wide location and function of DNA binding proteins. Science 290, 2306-
                          Kimura, H. and Cook, P. R. (2001). Kinetics of core histones in living human        2309.
                            cells: little exchange of H3 and H4 and some rapid exchange of H2B. J. Cell     Riddick, G. and Macara, I. G. (2005). A systems analysis of importin-{alpha}-
                            Biol. 153, 1341-1353.                                                             {beta} mediated nuclear protein import. J. Cell Biol. 168, 1027-1038.
                          Kimura, T., Mills, F. C., Allan, J. and Gould, H. (1983). Selective unfolding     Robert, F., Pokholok, D. K., Hannett, N. M., Rinaldi, N. J., Chandy, M.,
                            of erythroid chromatin in the region of the active beta-globin gene. Nature       Rolfe, A., Workman, J. L., Gifford, D. K. and Young, R. A. (2004).
                            306, 709-712.                                                                     Global position and recruitment of HATs and HDACs in the yeast genome.
                          Kurdistani, S. K., Tavazoie, S. and Grunstein, M. (2004). Mapping global            Mol. Cell 16, 199-209.
                            histone acetylation patterns to gene expression. Cell 117, 721-733.             Roh, T. Y., Cuddapah, S. and Zhao, K. (2005). Active chromatin domains
                          Lamond, A. I. and Earnshaw, W. C. (1998). Structure and function in the             are defined by acetylation islands revealed by genome-wide mapping. Genes
                            nucleus. Science 280, 547-553.                                                    Dev. 19, 542-552.
                          Lander, E. S., Linton, L. M., Birren, B., Nusbaum, C., Zody, M. C.,               Rout, M. P., Aitchison, J. D., Suprapto, A., Hjertaas, K., Zhao, Y. and
                            Baldwin, J., Devon, K., Dewar, K., Doyle, M., FitzHugh, W. et al. (2001).         Chait, B. T. (2000). The yeast nuclear pore complex: composition,
                            Initial sequencing and analysis of the human genome. Nature 409, 860-921.         architecture, and transport mechanism. J. Cell Biol. 148, 635-651.
                          Lee, C. K., Shibata, Y., Rao, B., Strahl, B. D. and Lieb, J. D. (2004).           Rout, M. P., Aitchison, J. D., Magnasco, M. O. and Chait, B. T. (2003).
                            Evidence for nucleosome depletion at active regulatory regions genome-            Virtual gating and nuclear transport: the hole picture. Trends Cell. Biol. 13,
                            wide. Nat. Genet. 36, 900-905.                                                    622-628.
                          Lee, R. C., Feinbaum, R. L. and Ambros, V. (1993). The C. elegans                 Saitoh, N., Spahr, C. S., Patterson, S. D., Bubulya, P., Neuwald, A. F. and
                            heterochronic gene lin-4 encodes small RNAs with antisense                        Spector, D. L. (2004). Proteomic analysis of interchromatin granule
                            complementarity to lin-14. Cell 75, 843-854.                                      clusters. Mol. Biol. Cell 15, 3876-3890.
                          Levens, D. L. (2003). Reconstructing MYC. Genes Dev. 17, 1071-1077.               Schena, M., Shalon, D., Davis, R. W. and Brown, P. O. (1995). Quantitative
                          Lever, M. A., Th’ng, J. P., Sun, X. and Hendzel, M. J. (2000). Rapid                monitoring of gene expression patterns with a complementary DNA
                            exchange of histone H1.1 on chromatin in living human cells. Nature 408,          microarray. Science 270, 467-470.
                            873-876.                                                                        Scherl, A., Coute, Y., Deon, C., Calle, A., Kindbeiter, K., Sanchez, J. C.,
                          4092       Journal of Cell Science 118 (18)
                            Greco, A., Hochstrasser, D. and Diaz, J. J. (2002). Functional proteomic      Tange, T. O., Nott, A. and Moore, M. J. (2004). The ever-increasing
                            analysis of human nucleolus. Mol. Biol. Cell 13, 4100-4109.                     complexities of the exon junction complex. Curr. Opin. Cell Biol. 16, 279-
                          Schirmer, E. C., Florens, L., Guan, T., Yates, J. R., 3rd. and Gerace, L.         284.
                            (2003). Nuclear membrane proteins with potential disease links found by       Venter, J. C., Adams, M. D., Myers, E. W., Li, P. W., Mural, R. J., Sutton,
                            subtractive proteomics. Science 301, 1380-1382.                                 G. G., Smith, H. O., Yandell, M., Evans, C. A., Holt, R. A. et al. (2001).
                          Schubeler, D., Scalzo, D., Kooperberg, C., van Steensel, B., Delrow, J. and       The sequence of the human genome. Science 291, 1304-1351.
                            Groudine, M. (2002). Genome-wide DNA replication profile for                   Waterston, R. H., Lindblad-Toh, K., Birney, E., Rogers, J., Abril, J. F.,
                            Drosophila melanogaster: a link between transcription and replication           Agarwal, P., Agarwala, R., Ainscough, R., Alexandersson, M., An, P. et
                            timing. Nat. Genet. 32, 438-442.                                                al. (2002). Initial sequencing and comparative analysis of the mouse
                          Schubeler, D., MacAlpine, D. M., Scalzo, D., Wirbelauer, C., Kooperberg,          genome. Nature 420, 520-562.
                            C., van Leeuwen, F., Gottschling, D. E., O’Neill, L. P., Turner, B. M.,       White, E. J., Emanuelsson, O., Scalzo, D., Royce, T., Kosak, S., Oakeley,
                            Delrow, J. et al. (2004). The histone modification pattern of active genes       E. J., Weissman, S., Gerstein, M., Groudine, M., Snyder, M. et al. (2004).
                            revealed through genome-wide chromatin analysis of a higher eukaryote.          DNA replication-timing analysis of human chromosome 22 at high
                            Genes Dev. 18, 1263-1271.                                                       resolution and different developmental states. Proc. Natl. Acad. Sci. USA
                          Shav-Tal, Y., Darzacq, X., Shenoy, S. M., Fusco, D., Janicki, S. M., Spector,     101, 17771-17776.
                            D. L. and Singer, R. H. (2004). Dynamics of single mRNPs in nuclei of         Wightman, B., Ha, I. and Ruvkun, G. (1993). Posttranscriptional regulation
                            living cells. Science 304, 1797-1800.                                           of the heterochronic gene lin-14 by lin-4 mediates temporal pattern
                          Simpson, J. C., Wellenreuther, R., Poustka, A., Pepperkok, R. and                 formation in C. elegans. Cell 75, 855-862.
                            Wiemann, S. (2000). Systematic subcellular localization of novel proteins     Woodfine, K., Fiegler, H., Beare, D. M., Collins, J. E., McCann, O. T.,
                            identified by large-scale cDNA sequencing. EMBO Rep. 1, 287-292.                 Young, B. D., Debernardi, S., Mott, R., Dunham, I. and Carter, N. P.
                          Slepchenko, B. M., Schaff, J. C., Carson, J. H. and Loew, L. M. (2002).           (2004). Replication timing of the human genome. Hum. Mol. Genet. 13,
                            Computational cell biology: spatiotemporal simulation of cellular events.       191-202.
                            Annu. Rev. Biophys. Biomol. Struct. 31, 423-441.                              Yuan, G. C., Liu, Y. J., Dion, M. F., Slack, M. D., Wu, L. F., Altschuler, S.
                          Slonim, D. K. (2002). From patterns to pathways: gene expression data             J. and Rando, O. J. (2005). Genome-scale identification of nucleosome
                            analysis comes of age. Nat. Genet. 32, 502-508.                                 positions in S. cerevisiae. Science 309, 626-630.
                          Smith, A. E., Slepchenko, B. M., Schaff, J. C., Loew, L. M. and Macara,         Yuh, C. H., Bolouri, H. and Davidson, E. H. (1998). Genomic cis-regulatory
                            I. G. (2002). Systems analysis of Ran transport. Science 295, 488-491.          logic: experimental and computational analysis of a sea urchin gene. Science
                          Spector, D. L. (2003). The dynamics of chromosome organization and gene           279, 1896-1902.
                            regulation. Annu. Rev. Biochem. 72, 573-608.                                  Yuh, C. H., Bolouri, H. and Davidson, E. H. (2001). Cis-regulatory logic in
                          Sprague, B. L. and McNally, J. G. (2005). FRAP analysis of binding: proper        the endo16 gene: switching from a specification to a differentiation mode
Journal of Cell Science

                            and fitting. Trends Cell. Biol. 15, 84-91.                                       of control. Development 128, 617-629.
                          Strahl, B. D. and Allis, C. D. (2000). The language of covalent histone         Zeitlinger, J., Simon, I., Harbison, C. T., Hannett, N. M., Volkert, T. L.,
                            modifications. Nature 403, 41-45.                                                Fink, G. R. and Young, R. A. (2003). Program-specific distribution of a
                          Tackett, A. J., Dilworth, D. J., Davey, M. J., O’Donnell, M., Aitchison, J.       transcription factor dependent on partner transcription factor and MAPK
                            D., Rout, M. P. and Chait, B. T. (2005). Proteomic and genomic                  signaling. Cell 113, 395-404.
                            characterization of chromatin complexes at a boundary. J. Cell Biol. 169,     Zhou, Z., Licklider, L. J., Gygi, S. P. and Reed, R. (2002). Comprehensive
                            35-47.                                                                          proteomic analysis of the human spliceosome. Nature 419, 182-185.

To top